CGeNArate

CGeNArate generates sequence-dependent coarse-grained models for DNA molecular dynamics simulations, enabling accurate atomistic resolution of kilobase-long B-DNA duplexes with reduced computational effort for biotechnological and chromatin studies.

Key Features:

• Sequence-dependent coarse-grained Hamiltonian: The model captures sequence-specific effects on DNA structure and dynamics.
• Machine Learning back-mapping: Coarse-grained trajectories are converted to atomistic resolution with extreme accuracy using Machine Learning Approaches.
• Global and local property reproduction: The method accurately reproduces both local and global physical properties of DNA.
• Kilobase-scale simulations: The efficiency enables atomic-resolution ensembles of DNA segments containing many kilobases, entering the gene range or encompassing entire organelle genomes.

Scientific Applications:

• Biotechnological studies: Simulations of very long B-DNA segments support biotechnology research applications.
• Chromatin studies: The method enables investigation of DNA at scales relevant to chromatin structure and function.

Methodology:

CGeNArate uses a coarse-grained Hamiltonian with trajectories back-mapped to atomistic resolution by means of Machine Learning Approaches.